Research Fellow
Core
Build open benchmarks, evaluation environments, and datasets to measure AI progress and capabilities.
Role type
Research Fellow (PhD/Postdoc)
Builds
Open-source academic artifacts (benchmarks, datasets, evaluation environments)
Domain
Artificial Intelligence / Frontier AI / Agentic AI
Deliverable
production ML models
Required skills
Research design, benchmark creation, dataset construction, code implementation, domain expertise, expert collaboration
Preferred skills
Publications at NeurIPS/ICML/ICLR, agentic evaluation, RL environments, LLM evaluation harnesses
Technologies
LLM evaluation harnesses, RL environments
Responsibilities
Scope and lead new benchmarks or datasets, define task specifications and rubrics with domain experts, run expert review processes, release research artifacts publicly, participate in weekly research meetings
Seniority
Mid-level, hands-on IC
